-1-

  The downstream photon veto counter(DPV) is designed  to detect 

photons from kaon decays and other radiative proceses in the beam 

direction through the target of the E949 Detector. This counter 

is able to detect minimum ionization particles (MIP) and photon 

and hadron showers.

  The DPV consists of 26 layers of 10mm thick scintillator alter-

nating with 25 layers of 1.5mm thick lead inside of a hard alumi-

nium cover. Each layer of scintillator consists of three scintil-

lation sheets of size 233mm x 700mm x 10mm with 45 degree overlap-

ping joints. Each scintillation sheet is wrapped with 0.02mm thick 

Al mylar film except the two faces opposite the PMT windows. The 

efficient area of the counter is 700mm x 700mm, the thickness in 

the beam direction is 310mm or 7.3 radiation lengths. The weight 

of the counter is 450kg including 200kg lead and 127kg scintillator. 

The design of the DPV is shown on the fig.1. There are two types 

of plastic scintillator in the DPV assembly: "extruded" (e) and 

"machined" (m). The layout of the scintillator-lead layers along 

the beam direction is shown on the fig.2. Extruded and machined 

scintillators have different attennuation lengths: 450mm for 

machined and 1050mm for extruded (Fig. 3) which was taken into 

account in the construction of the counter.
  
  Scintillation light is collected from scintillator ends by means 

of four reflectors and four photomultipliers (PMT-8854). The light 

reflection surfaces of the reflectors are made of aluminized mylar. 

The scheme of the reflector - PMT-8854 joint is shown on the fig.4. 

The main features of PMT-8854 are the following: photocathode dia-

meter is 114mm; 14 stages of electron multiplication; UV response 

to 220nm; quantum efficiency 22.5% at 385nm; electron transit time 

66ns at 3000v. Each PMT is protected from 

                            -2-

the fringe magnetic field by a special steel cover. The schematic  

diagram of the PMT-8854 base is shown in the fig.5. 

  The DPV is instrumented with one blue LED NSPB300A from NICHIA 

corporation. The LED is installed in the center of last (downstream)

scintillation layer and provides light to all PMTs simultaneously. 

The LED is driven by pulses of positive polarity with amplitude in 

the range (1.5 - 3.5)V and width of 20ns. The  LED allows tracking 

PMTs' gain instabilities and rapid calibration.

  For HV adjustment and for control of the LED calibration each PMT has 

a YAlO3:Ce-Am241 light pulser mounted on the PMT window by means of sili-

con grease. The typical response of the PMT-8854 to the YAlO3:Ce-Am241 

light pulser is shown in fig.6. A "cookie" was used to attennuate the 

light yield of each YAlO3:Ce-Am241 to the dynamic range of the ADC. 

The spectra of YAlO3:Ce-Am241 in real DPV geometry for all PMTs are shown 

in fig.7a-e. The mean amplitudes (with "cookie") of each YAlO3:Ce-Am241 

response are 149,97,115,106pC respectively. The FWHM - 8% and it corres-

ponds to Np.e.- 800 for each pulser. There is a small muon background at 

low yield in the YAlO3:Ce-Am241 spectrum. The frequency of the YAlO3:Ce-

Am241 light pulser is about 40Hz; sizes: 5mm diameter and 1mm height.

  The DPV was tested with cosmic muons. Two trigger scintillation 

counters of size 10cm x 11cm were used to produce a muon trigger.  

Fig.8 shows total response of all 4 PMTs as a function of the muon 

location (along scintillator sheet axis) crossing all layers of the 

counter. It is a calculated total PMT response for DPV from measure-

ment of one end of each scintillator type to determine their attennua-

tion lengths (see fig.3). On this figure are also indicated  experi-

mental results for minimal and maximal light output from real scintil-

lator geometry. These results coincide with the calculated light out-

put from scintillator. The light yield in the middle plane of the DPV

                            -3-

(center of scintillation sheets) is uniform. So the minimal and maxi-

mal light output from DPV scintillator consists of 9.1p.e./Mev and 

12.1p.e./Mev respectively. The uniformity of the light yield along the 

scintillator sheet length is 74.8%.

  To illustrate the DPV efficiency figs.9a-e show amplitude spectra 

for PMT#1,2,3,4 and their analog sum for muons crossing the center 

of the DPV. Each PMT was calibrated using the correlations between LED 

mean amplitudes and light outputs determined from width. To determine 

the number of photoelectrons (Np.e.) a Gaussian fit was made to the 

LED spectra with large statistics - 10000 for each point. A typical 

LED amplitude distribution is shown on the fig.10. Correlations of 

the mean LED amplitude and light yield are shown on the fig.11.

  The DPV has high enough average light output - 10p.e./Mev  and is 

able to improve efficiency of the E949 Detector Veto system by incre-

asing the number of radiation length in the downsstream direction 

from 7 to 14.